Bone marrow transplantation (BMT) following ablation of the patient's hematapoietic system is a life-saving treatment for several types of leukemia. Acute graft vs. host disease (GVHD) is a serious complication in more than a third of bone marrow transplants (BMTs) between MHC matched donors and recipients, and the proximate cause of death in ~ 15% of BMT patients. Purging infused marrow of all donor T cells largely eliminates GVHD, but results in more frequent stem cell failure to engraft, and higher rates of leukemia relapse. The inverse correlation between GVHD and relapse is attributable to the graft vs. leukemia (GVL) response, whereby donor T cells recognize host alloantigens on residual leukemia cells. It is unclear whether sufficient numbers of donor T cells recognize leukemia specific antigens to afford protection in the absence of alloreactivity, but this issue is presently moot, since a technique for efficiently removing all alloreactive cells while retaining leukemia specific cells is not yet available. The severity of GVHD depends on the strength of the alloreactive response, the ability of donor T cells to infiltrate target organs (skin, gut, liver, lungs), and the susceptibility of those tissues to immune mediated damage. Given the problems inherent in eliminating alloreactive donor cells, the optimal theoretical approach might be to limit their access to host sites most susceptible to immunopathology, and/or most likely to induce reactivity. Our approach may accomplish this by inhibition of Rho associated coiled coil kinases (ROCK1&2) enzymes crucial for cytoskeletal function during leukocyte motility and loosening of endothelial and epithelial tight junctions. Three small molecule ROCK inhibitors will be tested in MHC and minor histocompatibility antigen (miHA) disparate BMT mouse models. Two inhibitors of ROCK1&2 (Telmisartan and Fasudil) have extensive clinical safety records as vasodilators. The third, SLx2119 is ROCK2 specific, and has just passed Phase 1 clinical trials on the path to development as an anti-cancer drug. Our preliminary data indicate that 10 days (but not 1 day) of Fasudil is highly protective against acute MHC disparate GVHD, without inducing allotolerance that could compromise beneficial GVL. We propose to extend this observation by testing the three mentioned drugs in both MHC and miHA models in 1 vs. 2 week schedules, to find the shortest duration of treatment that maximally suppresses GVHD. The optimal schedule for each drug will be used to suppress GVHD in the presence of infused host- syngeneic leukemia cells, and tumor burden will be monitored in treated and untreated mice. We hypothesize that ROCK inhibition will interfere with donor T cell alloactivation and/or trafficking to primary sites of GVHD, and that interference with alloactivation will not hinder donor T cells from finding and attacking more accessible leukemia cells. Our objective is to determine at least one schedule of ROCK inhibitor that suppresses GVHD while sparing GVL function. Using ROCK inhibitors for GVHD is novel, and success would accelerate our long term goal of using ROCK inhibitors with BMT, to treat leukemias and other tumors.